Web Image Re-Ranking Using Hash Based Signatures

This project addresses content-based image retrieval in general, and in particular, focuses on developing a hidden class detection methodology to address effective semantics-intensive image retrieval. In our approach, each image in the database is segmented into classes and contains classified images. We explore the query adaptive ranking to retrieve images. With this representation, model based on hash code of the image database is obtained, to analyze semantic concepts in the database. The semantic similarity is measured through sum rates for detecting images, to the discovered semantic signatures. And also implement the Hamming distance techniques to retrieve the relevant images from databases. The proposed schemes were compared with a conventional CBIR scheme employing image classification. Our experimental results are providing good retrieval on re-ranking precisions compared with the state-of-the-art methods.